An engagement heatmap is a visualisation that shows, for each moment in a video, what fraction of the audience was watching, re-watching, or had already dropped off, typically rendered as a colour gradient laid over the video timeline. The underlying data comes from playback telemetry — played, paused, seeked, and completed events from the xAPI Video Profile — aggregated across all learners who have watched the asset and stored in an LRS (Learning Record Store). Segments shown in a "hot" colour (many viewers present) indicate content that draws and holds attention; "cold" segments reveal where learners quit or skip forward; spikes of re-watch highlight moments of high perceived importance or confusion that warranted a second pass. For instructional designers the heatmap is the most actionable video quality signal available: it turns an abstract completion rate into a precise editorial cue — "the explanation at 4:23 is where everyone drops off, refilm it". A practical limitation is that heatmaps require a large enough audience to be statistically meaningful; on a cohort of ten learners the pattern is too noisy to act on. Combining heatmap data with in-video quiz performance at the same timestamps creates a richer picture: a cold segment that also correlates with a low quiz score on that concept is a definitive signal that the explanation failed. Unlike aggregate metrics, a heatmap preserves the temporal dimension of learning behaviour, making it uniquely suited for driving precise content revisions rather than wholesale re-recordings.
